The Happy Vacuum Cleaner

One of several autonomous vacuum cleaners - the Kobold VR100

Kobold VR100

In 2005, I organized a symposium at the 14th Conference of the International Society for Research on Emotions (ISRE), which was held in July (11-15) in Bari Italy. In this context I gave a talk, which I had given the title MY HAPPY VACUUM CLEANER.

If I remember correctly, a considerable number of my dear colleagues initially stared at me as if I was nuts. The talk was a conceptual piece that traced issues in creating robots as helpers or companions from a point of view of affective processes and the complexity of the role such artificial entities might take on for the social and affective life of their humans. I started the presentation with reading a short piece of (my own) fiction – basically a no-no at a serious scientific conference. Having said that, ISRE is a rather special group of people that agree that emotions are processes of such complexity that it requires serious interdisciplinarity – and this bridges not only the social, natural, and behavioral sciences, but also engineering and the humanities. In that spirit I jumped into the cold water and read the vacuum diaries – which are here republished for the first time:

Arvid Kappas – The Vacuum Cleaner Diaries (2005)

Tuesday: I return late from work and find a note – the package has finally arrived! It has been three weeks that I waited for my new VacBot. The post man left the box with the neighbors. It is too late to try everything out – I guess I will just open the box and charge my new high tech gadget, a RoboVac 2400 intelligent vacuum cleaner with EmoTronics. The cat throws a skeptical look at the box! The robot is a shiny red disk with wheels at the bottom, diameter that of a large pizza. Oh yes, Pizza. I must get something to eat …

Wednesday: Today is the day … I had started reading the instruction documents, of which there were very few. The RoboVac has all of the documentation built in and negotiated installation of all software with the wireless house net – so some of it actually lives in my house PCs. First came the general greeting session where RoboVac gets to know me, what I look like and what I sound like. In turn I am greeted with a series of happy and not so happy bleeping and whistling to familiarize myself with the basic messages. As far as I remember, the company did some pretests and the last generation of talking VacBots was a major flop. OK – the setup is complete and the info sounds all rather clear, I guess it is off for the ‘bot to go and explore the apartment …

Thursday: I must say, it was really fun to watch the little thing explore the place. Cleaning should start in earnest today. When I came home today, the little guy came out of the kitchen and upon seeing me spun a few times around shooting off a cheerful set of bleeps and beeps and then returned to his work. Fantastic! I could not resist spying a little bit. Oh yes, that soup spot on the kitchen floor. Up and down, left and right he goes. This is determination I would say.  I remember that part of the EmoTronics set-up is a basic set of motivations and emotions – for example, if something is particularly difficult to clean, it is particularly rewarding to get rid of the dirt – another powerful reinforcer is if I praise him. “Well done VacBot! Well done! … of course a little dance and song and he continues. I must say that I start to enjoy the occasional beeping in the background underlining success in the fight against some dust bunnies while I correct some student papers in my office.

Friday: Odd. The little guy was spending some time again with the soup spot in the kitchen which seems to have long disappeared. I believe this is where I praised him. Mmmh.  “VacBot!” I tell the spinning disc “Well done – now move on” – of course, he responds with the sing and dance. I wonder how the cat is dealing with all of this. She has been a bit quiet in the last few days.

I also saw that if a place is associated with some problems, the VacBot will be very careful – in the corner over there I had leaned an umbrella which had fallen over when the ‘bot bumped against it … now I can see that this is a corner that is negotiated with utmost care – even if I have taken the umbrella away. Anyway … now he is recharging at his station.

Saturday: Today I will spend some more time with my new gadget. Where is it? Curious! “VacBot!! Identify your location!” I yell. A low beep bleep beep is coming out of the bathroom. What is going on? Did the wheels get stuck? When I checked everything out I saw that the little guy was repeatedly approaching the door from the inside, but turned away every time approaching the threshold. Interesting. I swear – the pattern looks like an approach – avoidance conflict. I guess I have to have a look at the diagnostics. After breakfast however! Ah, upon hearing me dropping corn flakes on the floor he makes it out of the bathroom. I will check the diagnostics in any case.

*** I can’t believe it! After trying all of the diagnostics and not finding a fault, I checked the surveillance video the VacBot constantly records and sends to the house network. I have found the problem. At 4.32 last night the ‘bot was attacked by my cat. Beaten up so to say. So that was the problem. The little machine was scared to leave the bathroom. This is silly. I understand that it is functional for my vacuum cleaner to be careful when something … ahem … dangerous could happen, but this is silly. And now. No. He is in the kitchen again – the soup spot. I guess I should have been careful with my praises. Argh. And yet. When he sees me he does a little dance routine and beeps and whistles a little tune. Do you think he is maybe relieved because in my presence the cat won’t attack?

Sunday: After some more interesting side effects of the EmoTronics which would take too long now to write down I have come to the conclusion that I am better off resetting the software to the previous version. Autonomous, Intelligent. But no Emotions! Who needs a vacuum cleaner that is afraid of your cat?

Monday: This is the end of spending my first week with the VacBot and I got used to him. No sing and dance when I get home now. No rushing into the kitchen to get praise for the non-existent soup spot. Sad. Sort of.

Tuesday: I have given up and reactivated the EmoTronics. We watched a movie tonight together. Terminator V! I know that Marvin – I had to give my ‘bot  a name after all – does not enjoy the movie itself, but he likes to be near me and I drop some popcorn on the floor from time to time which he sucks up because he likes it clean and he likes to please me. “Good boy. Good boy.” He utters some happy bleeps and beeps. The cat is sitting resigned on my lap. I am smiling.

Cover of "Wall-E (Single-Disc Edition)"

Cover of Wall-E (Single-Disc Edition)

In 2005, Roombas and similar devices only started to appear, but had not yet reached a large degree of public attention. But of course, the talk was not about little disc-shaped vacuum cleaners, but a complex tale of empathy and interaction. It was important for me to underline that intelligent artificial companions would not need to be humanoid. They would not even need a face, such as Pixar’s Wall-E (2008). As I said in the Nautilus Interview (see previous post)

It is not surprising that we have empathy to robots. Everybody understands that they might react strongly to R2-D2 being trashed with a sledgehammer.

One of the key research issues that my group and many colleagues in the field of affective computing are trying to illuminate today is, what are the minimal conditions for human attachment to artificial entities? In my case I am interested in this question because I want to understand how humans link to each other and why and how they interact with each other. This is important, because demonstrating that very simple system properties give us the sense that the other is sentient and links with us in some way has the capacity to change the way we look at human interaction. I will give an example from social psychology: One of the earliest topics in social psychology was the study of the phenomenon of social facilitation. The fact that in the presence of others many behaviors are accelerated or amplified. To explain the effects that had been demonstrated in many empirical studies, complex theories were developed. However, in an attempt to demonstrate that the explanation might not be so complicated after all, the famous social psychologist Robert Zajonc demonstrated in 1965 related social effects with cockroaches (!). Of course, if cockroaches show social phenomena, it is not necessary to use much brain to compare yourself with representations of others, in the light of public expectations yada yada yada. While the experiment and its interpretation are not without critics, it is resarch like this that causes a shift in thinking how we can conceive of social phenomena. In a similar way, it might be that the feeling of being connected with a human is not due to complex cognitive processes, but a handful of simple features, such as behavioral synchronicity. These were the candidates I touched upon in the vacuum cleaner talk:

Candidates for minimal requirements for human attachment to robots (Kappas, 2005)

  • Morphological features?
  • Emotional expressions?
  • Intelligence?
  • Reactivity?
  • Synchronicity?

What makes this question so interesting is that there are issues that complicate the matter – in procedural terms, they interact with the features of the artificial systems, such as properties of the human (age, socio-cultural aspects of how they were brought up), experience, and of course the specific situational context. These are psychological aspects that make life complicated for the engineer who wants to design a machine and they make life interesting for the behavioral scientist.


Kappas. A. (2005). My happy vacuum cleaner. Presented at the XIVth Conference of the International Society for Research on Emotions, Bari, Italy (July 2005).

Zajonc, R.B., Heingartner, A., & Herman, E.M. (1969). Social enhancement and impairment of performance in the cockroach. Journal of Personality and Social Psychology, 13, 83-92.


Artificial Emotions


Nautilus (Photo credit: Lebatihem)

Artificial Emotions

Today an intriguing new magazine saw the light of day: Nautilus.

Nautilus is a different kind of science magazine. We deliver big-picture science by reporting on a single monthly topic from multiple perspectives.

The first issue poses the question What makes you so special? It deals with what it might mean to be human — and how we might or might not be unique. One of the stories in the first issue deals with Artificial Emotions – As it happens, I had a long conversation with the author of the story, Neil Savage. Some of that ended up in the article, among many quotes by some of my colleagues, reflecting on issues central to affective computing. And this is why it fits wonderfully with the topic of this blog.


Nautilus (Photo credit: Guilli F P)

Every time I discuss issues of artificial emotions with people, regardless of whether they are scientists, represent the media, or others who are simply curious about affective computing, there are certain topics that tend to come up again and again. One of my goals is to raise some of these over the course of the next weeks in this blog. For example

  • that we might require much less emotion than we think to feel that machines are/feel emotional.
  • How psychological theories might be misleading efforts to make machines emotion-savvy.
  • How the concept of emotions as such might not always be very useful at all in this context.

In the meanwhile – check out Neil’s article, it is an interesting read .— and I will check out the rest of Nautilus in the meanwhile